Confirmation of Payment to Shapeways

It should be delivered to me in 10 days! :)

Shapeways Models Uploaded and Available

Editing Models In 3DS Max 2011

It is hard to see in these screenshots, but I have added a small sphere in the middle of each of these models (they are blue and purple) because at the centre of the series of circles that I created in Grasshopper the pipes did not connect properly and made the model look very untidy. And I added some spheres at some of the intersections as well to stop there from being holes in the corners.

I don't know why the models came out in 2 different colours, but it didn't matter when exporting it as a .stl file for 3D printing anyway.

Baked Models Screenshots

These models were baked in Grasshopper into Rhino, then scaled in Rhino itself by 10. This is because my original scale was so small in Grasshopper, so this made it so that the pipes are 1mm thick, which I think is reasonable and will give me the aesthetics I want, and the diameter of the models will be around 15mm.

To make the bottom of the models neater I found the bottom circle that made up the dome, and piped it. This made the model much neater altogether and the fabrication will work well with this included in the model.

You can see in each of these models you can see some imperfections in some of the intersections, like parts of the pipes sticking out further past the circle surrounding the model, and some of the intersections between pipes are not neat enough either.

So I imported the models into 3DS Max to edit the mesh and add spheres to clean up intersections.

The Triangle-Based Model, What Went Wrong

This is a screenshot of the triangle model in Rhino after it was baked in Grasshopper and then scaled by 10 in Rhino, with the viewing setting at shaded. You can see that parts of the pipes do not go together well and intersect with overlaps. I was hoping to fix this in 3DS Max.


Here is the imported model to 3DS Max, exported as a .3ds file from Grasshopper. It seemed to import very strangely into 3DS because of the 2 different colours of elements in the model, seeming to be randomly different, but it didn't seem to make a difference when editing the model anyway. It is hard to see in this model, so I have circled some of the problem areas in the image below, but if you look closely at this image you can see that it is happening at almost every intersection of pipes in the model. I was finding that if I 'edit mesh' and select some points at the end to delete, the pipes would not be long enough to still join together, so then I decided to try scaling them, but then they scale from the middle of the element, and to get it back to the right position I had to move it, which made it hard to get the model exactly as I exported in the first place.

This was a very annoying process and the model did not look as even in the pattern as I wanted it too, so decided to go with a different one.

3 Iterations of Model Chosen

This iteration has very small holes in it because in the Grasshopper file the size of the polygons is large and the grid size is small (in relative terms), which means that more of the edges of each polygon intersect with one another. For this model the 'control points' in the Grasshopper file were mainly the grid size (1mm in this case) and the polygon size (1.5mm) which are then scaled by 10 for fabrication. The model is so dense because both the values for these points are low, so everything is very close together.

This iteration is more low density than the first, because the grid size and polygon size are more similar (1.9mm grid and 2.5mm polygon size- then scaled by 10 for fabrication). This medium density should create some interesting shadows because there are a few different sized and shaped voids in the model.

The one above I wanted to use because my polygons were triangles instead of octagons like the other iterations, but I found when I baked the model that there were all sorts of problems with almost every intersection of pipes, and it would be very hard to fix even in 3DS Max like I had been doing with the others so I decided to use a different iteration.

This iteration is being used because more of the polygons intersect but not as much as the first model which makes a very unexpected and interesting pattern (from octagons) on the surface of the dome. The grid size for this iteration was 1.9mm and the polygons were 3mm- then scaled by 10 for fabrication after baking)

First Successful Upload to Shapeways!

http://www.shapeways.com/model/264954/bec_model_stl_origin.html?gid=ug85099

This was a test model to the right scale to see if there are any errors in my model that would stop it from being uploaded before I spend time perfecting the geometry for the real model I want to print. For this model I baked the joined geometry in Grasshopper, and then scaled my model by 10 in Rhino, which made the pipes about 1mm in diameter which is printable.
This is my 4th attempt at uploading a model to this website and this one did not have an error! There are a couple of mistakes in the model, you can see them in this image so now I am going to figure out how to edit the baked model in Rhino to make it exactly what I want for 3D printing! It's about 14cm in diameter and will be about $20 for each iteration that I print :)

Here is the email they sent me when the model was successfully uploaded and checked for errors!


Here is my Grasshopper model and what it looked like before being uploaded. The red is the Grasshopper-generated model, and the green is the baked and scaled version of that that I uploaded.

3D Printing

Josh Harle had the idea that I should get my model 3D printed instead of laser cut. Here are some videos that I found on the website. This is the material I think would be really good, and I found them really interesting!





http://www.shapeways.com/

Josh Harle in Lab Class

This week (week 11) I asked Josh Harle to help me with the model again, because I did not understand the changes he had made to the earlier version, and I found out that he thought the model looked different when he finished with it. So we worked together in the tutorial to make the changes I was originally after and here are the results:




I would still like to close the gap in the middle, but it is not essential, but I would like to make the top a bit more neat somehow, maybe make a rule that says the ends should meet up with the closest other end? I will have a think about this.

New Square Grid

After looking at that last file, I figured out how to change what was projected to just a normal square grid that fits over the dome to project these shapes. This gave the more organised and consistent affect I was after. Here is an image of how it works in Grasshopper and Rhino.

Here are some images of the different iterations of this system. I really like it but I wonder how I would get it laser cut. I am planning on dividing up the dome into triangles and gluing them together to create the dome shape, but with this pattern there are no solid areas that could be glued together when it is cut up.
I will have to think of ways to do this a better way, or change the look of my model in some way to accommodate for the limits of laser cutting.

Josh Harle's Help!

I had a look at changing my system to do what I wanted, but I was having a lot of trouble with it so decided to see if Josh Harle could help from the forum on Blackboard. Here are some screenshots of what Josh gave back to me :)


I didn't see much of a difference but the hole in the top was closed which is good! I actually found out later that it worked on Josh's computer and he was very confused that this is the file that I got back!
When I got this back, I thought that maybe I would prefer it if I started with a completely different grid to project from rather than the series of circles.

Testing In Grasshopper

Here I changed the shape I used to project onto the surface from circles to hexagons. I really prefer this because it looks more geometric and it looks more like The Louvre design (structural image in another post). I also really like the different intersections of shapes I am getting from this shape, and the new shapes that are created from this. I am looking forward to exploring this more.
In these 3 iterations all I have changed is the size of the polygon that is being projected- there are the same number of polygons with the same amount of sides.
I would like to find a way to change this so that there is not a big hole in the top of the dome, and I don't like that the shapes become more dense as you get closer to the top of the dome. This happens because each circle in the series is divided into the same number of points, so points are closer together on the smaller circle at the top than the bottom one. I would also really like the pattern to be less random-looking and look more organised and more consistent.

Example I Found

Here is an example of a pattern made in Grasshopper which I think is very cool! I wouldn't be able to make this system myself, (it was barely worth showing the system in this screenshot because you can't even see the nodes! but I wanted to show how complex the system was!) but I would like to make a similar pattern for the surface of my dome, hopefully using less nodes.

File (hexagon_wave.ghx) downloaded from a post by Liam Morrow, September 24th 2010 at 4.26pm. http://www.grasshopper3d.com/forum/topics/distort-a-hexagon-pattern. Retrieved April 28th 2011.

Modelling Proposal- What I submitted

The chosen site (image below taken from Project Brief, Jeremy Harkins, UNSW) for this proposal is the site of the Bennelong Apartments by Andrew Andersons and PTW Architects, commonly known as The Toaster. The current building on site is thought of as out of character for the harbour-side area and the brief calls for a structure that enhances the nature of the area. The site requires apartments, restraints/cafes and some interaction with the parkland and harbour, which means that pedestrian through fare is very important for this site, particularly as it is a popular area for tourists.

The proposed form of the project will largely focus on including natural light and airflow within a clearly defined area where the types of spaces required can share the benefits of them. Perhaps there is no practical reason for having natural light in a building design except in order to satisfy the human spirit, but the whole point of a piece of architecture is to induce a certain mood into the inhabitant, and keep us in touch with the flow of nature while empowering us to create our own spaces. A building that will embody this ideal is The Louvre in Abu Dhabi.

Fig.1 shows what the outside of The Louvre might look like when the project is finished. The form is a definite dome shape that lets a very controlled amount of light through to the ‘inside’ of the building. For the harbour-side site, this idea will be included in the design to give the area a certain amount of separation from the rest of the area surrounding, however the area will still feel like an outdoor area when standing underneath the dome structure, much more than what appears in Fig.2.

The proposed site is at the waterside just as this one in Abu Dhabi is, as the play of light on the waterfront creates a serene atmosphere that is not achievable in any other way. The building will take form as a part-undercover area with many different uses and facilities as separate buildings inside, including apartments, cafes and walkways to the parkland behind, and the focus on natural light is aimed to give the area fluidity and a relaxed mood for such a popular area.

While the design will be largely based on the design of The Louvre in Abu Dhabi, the concept of emergence will be considered when forming the structure which means that control points will be found in the system that can alter the design without losing control over the aspect of natural light for the area. When thinking about how this ‘roof’ structure might work this building was looked at again, and a system very similar to this will be used (Fig.3). The Sydney climate has also been considered, in that it is similar to Abu Dhabi in summer, but in winter the temperature drops dramatically and we get wind and rainfall, however under the dome is still to be considered outside space, just slightly undercover, and the individual businesses can account for the weather changes in their own way.

Fig.4 below shows the start on using Grasshopper for this project and finding the points of control, a concept to do with emergence, and here I have changed the radius of the circles being projected onto the dome shape and therefore created different densities of piped lines. These four forms look similar to the one above (Fig.3) but circles have been used and projected onto the dome shape, and perhaps it would be better to have shapes that are more geometrical and can intersect in interesting ways. This will be explored with further experimentation in order to achieve the required effect of the model. This is obviously only the ‘roof’ structure, and the buildings that will need to be placed below will also be designed and will essentially hold up this main attraction of the site.

Image References:

Fig 1: Jane and Mark Burry, The New Mathematics of Architecture (London: Thames & Hudson, 2010), 62.

Fig 2: Jane and Mark Burry, The New Mathematics of Architecture (London: Thames & Hudson, 2010), 62.

Fig 3: Jane and Mark Burry, The New Mathematics of Architecture (London: Thames & Hudson, 2010), 64.

Fig 4: Captures from author's computer, 10th May 2011.